The effects of the historical practice of residential redlining in the United States on recent temporal trends of air pollution near New York City schools

Background: In the 1930's the United States (US) sponsored Home Owners’ Loan Corporation (HOLC) created maps that determined risk for mortgage lending based on the racial and ethnic composition of neighborhoods leading to disinvestment in “redlined” or highest risk neighborhoods. This historical practice has perpetuated racial and economic segregation, and health disparities, that persist today. Interventions near schools where children spend large portions of the day, could impact large groups of children but schools are an often-overlooked environment for exposure. Despite a declining trend of ambient pollution in New York City (NYC) between 1998 and 2012, little is known about differences in air quality improvement near schools by historical redlining neighborhood status. Our objective was to examine if recent temporal trends of air pollution near NYC public schools differed in historically redlined neighborhoods. Methods: We examined annual average street-level concentrations of combustion-related air pollutants (black carbon (BC), particulate matter (PM_2.5), nitrogen dioxide (NO₂), and nitric oxide (NO)), within a 250-m radius around schools using NYC Community Air Survey land-use regression models (n = 1,462). Year of monitoring, historical redlining (binary), and summer ozone were included in multivariable linear regression using generalized estimating equation models. Average annual percent change (APC) in pollutant concentration was calculated. Models were further stratified by historical redlining and a multiplicative interaction term (year of monitoring × historical redlining) was used to assess effect modification. Results: Overall, there was a decreasing trend of BC (APC = -4.40%), PM_2.5 (-3.92%), NO₂ (-2.76%), and NO (-6.20%) during the 10-year period. A smaller reduction of BC, PM_2.5 and NO was observed in redlined neighborhoods (n = 722), compared to others (n = 740): BC (APC: −4.11% vs −4.69%; P_interaction < 0.01), PM_2.5 (-3.82% vs −4.11%; P_interaction < 0.01), and NO (-5.73% vs −6.67%; P_interaction < 0.01). Temporal trends of NO₂ did not differ by historical redlining (P_interaction = 0.60). Conclusions: Despite significant reductions in annual average pollution concentrations across NYC, schools in historically redlined neighborhoods, compared to others, experienced smaller decrease in pollution, highlighting a potential ongoing ramification of the discriminatory practice.